TREND-14 · UNG

Linear Regression Channel · Strategy evidence

Inspect the inputs, parameters and accounting behind a published result.

Run reference50e579e8269b

Last included price bar2026-11-10

Bundle generated (UTC)2026-11-11 10:55 UTC

Accounting modelSynthetic daily target allocations · v3

Replay the recorded returns and equity with the files below. Recreating strategy decisions needs the original matching inputs; fresh price downloads may differ. These are simulated allocations, not broker fills.

Downloads

Downloads load and verify compressed objects only when selected. Negative weights represent short exposure; no code is executed in your browser.

1. Download a complete bundle

Download the Python tool above and run this command. It checks shared objects and creates replay.zip; no repository clone is needed.

python e0543584b55263609cc8fd38cd032cb8432af6c2b48da3f53599e6e68c922f59.py https://quantarenas.com/data/evidence/runs/50e579e8269be60456602302f86093b53f5a752926091fe6539db646ced04371/manifest.json --bundle replay.zip

2. Replay the arithmetic

Extract replay.zip and run from its folder with Python. This verifies recorded weights, contributions and equity using only the standard library; it does not recalculate strategy decisions.

python scripts/replay_evidence.py data/evidence/runs/50e579e8269be60456602302f86093b53f5a752926091fe6539db646ced04371/manifest.json

3. Obtain and compare inputs

In the extracted bundle, use the recorded Python version and install the pinned dependencies. This command downloads provider inputs privately and reports inputHashMatches. Only a matching hash supports exact recomputation; revised provider data may not match. You may instead supply an original matching snapshot as private-input.json.

python -m pip install -r requirements-lock.txt
python scripts/recompute_evidence.py data/evidence/runs/50e579e8269be60456602302f86093b53f5a752926091fe6539db646ced04371/manifest.json --fetch-input private-input.json

4. Recompute the original strategy

With the exact original input snapshot and recorded dependencies, run the archived strategy and walk-forward selection. Original provider inputs are not public. Revised downloads fail the input-hash match and cannot establish exact reproduction.

python scripts/recompute_evidence.py data/evidence/runs/50e579e8269be60456602302f86093b53f5a752926091fe6539db646ced04371/manifest.json --input private-input.json

Accounting assumptions

Each day's return uses the preceding day's target allocation. The new target is decided at the current close. This is a synthetic close-based model, not a measured execution price.

Cash residual: 1 - sum(weights); residual may exceed one or be negative for shorts/leverage.
Decision: Target weights use bars through close t; previous targets earn (t-1,t].
Trading costs (bp): 0
Execution assumption: Synthetic daily target allocations, not broker orders or fills; financing/borrow costs omitted.
Synthetic rebalance turnover: Sum(abs(target[t] - target[t-1]*(1+assetReturn[t])/(1+grossReturn[t]))); one-way risky notional at close t after returns, divided by pre-trade equity.
Return interval: Sum of previous target weight times each asset close-to-close return; cash return is zero.
Target weight change: Sum(abs(target[t] - target[t-1])); changes in rule targets only.
Walk-forward transition: Candidate retains its own hypothetical state; first test return uses previous published target.

Parameters used

{
  "active": {
    "period": 50,
    "std_dev": 2.4
  },
  "defaults": {
    "period": 50,
    "std_dev": 2.0
  }
}

Walk-forward selections

Default/pre-walk-forward history is not out-of-sample evidence. Each fold records its training and test boundaries; the selected candidate retains its own hypothetical state.

1 · 2013-01-04T00:00:00 → 2013-07-03T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": -0.17532398253143874,
  "run": 0,
  "testEnd": "2013-07-03T00:00:00",
  "testStart": "2013-01-04T00:00:00",
  "trainEnd": "2013-01-03T00:00:00",
  "trainStart": "2010-01-04T00:00:00"
}

2 · 2013-07-05T00:00:00 → 2014-01-03T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": -0.43181435144153374,
  "run": 0,
  "testEnd": "2014-01-03T00:00:00",
  "testStart": "2013-07-05T00:00:00",
  "trainEnd": "2013-07-03T00:00:00",
  "trainStart": "2010-07-06T00:00:00"
}

3 · 2014-01-06T00:00:00 → 2014-07-03T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": -0.3384366957755518,
  "run": 0,
  "testEnd": "2014-07-03T00:00:00",
  "testStart": "2014-01-06T00:00:00",
  "trainEnd": "2014-01-03T00:00:00",
  "trainStart": "2011-01-06T00:00:00"
}

4 · 2014-07-07T00:00:00 → 2015-01-07T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": -0.3236665274595545,
  "run": 0,
  "testEnd": "2015-01-07T00:00:00",
  "testStart": "2014-07-07T00:00:00",
  "trainEnd": "2014-07-03T00:00:00",
  "trainStart": "2011-07-07T00:00:00"
}

5 · 2015-01-08T00:00:00 → 2015-07-08T00:00:00

{
  "config": {
    "period": 60,
    "std_dev": 2.0
  },
  "isSharpe": 0.27996113332508465,
  "run": 7,
  "testEnd": "2015-07-08T00:00:00",
  "testStart": "2015-01-08T00:00:00",
  "trainEnd": "2015-01-07T00:00:00",
  "trainStart": "2012-01-09T00:00:00"
}

6 · 2015-07-09T00:00:00 → 2016-01-08T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": 0.039762350102033536,
  "run": 0,
  "testEnd": "2016-01-08T00:00:00",
  "testStart": "2015-07-09T00:00:00",
  "trainEnd": "2015-07-08T00:00:00",
  "trainStart": "2012-07-09T00:00:00"
}

7 · 2016-01-11T00:00:00 → 2016-07-11T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": -0.30411992020613066,
  "run": 0,
  "testEnd": "2016-07-11T00:00:00",
  "testStart": "2016-01-11T00:00:00",
  "trainEnd": "2016-01-08T00:00:00",
  "trainStart": "2013-01-11T00:00:00"
}

8 · 2016-07-12T00:00:00 → 2017-01-12T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": -0.12546757504843323,
  "run": 0,
  "testEnd": "2017-01-12T00:00:00",
  "testStart": "2016-07-12T00:00:00",
  "trainEnd": "2016-07-11T00:00:00",
  "trainStart": "2013-07-12T00:00:00"
}

9 · 2017-01-13T00:00:00 → 2017-07-13T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": -0.31011569183854026,
  "run": 0,
  "testEnd": "2017-07-13T00:00:00",
  "testStart": "2017-01-13T00:00:00",
  "trainEnd": "2017-01-12T00:00:00",
  "trainStart": "2014-01-13T00:00:00"
}

10 · 2017-07-14T00:00:00 → 2018-01-12T00:00:00

{
  "config": {
    "period": 50,
    "std_dev": 1.6
  },
  "isSharpe": 0.3038267604363141,
  "run": 3,
  "testEnd": "2018-01-12T00:00:00",
  "testStart": "2017-07-14T00:00:00",
  "trainEnd": "2017-07-13T00:00:00",
  "trainStart": "2014-07-14T00:00:00"
}

11 · 2018-01-16T00:00:00 → 2018-07-16T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": -0.690576647400336,
  "run": 0,
  "testEnd": "2018-07-16T00:00:00",
  "testStart": "2018-01-16T00:00:00",
  "trainEnd": "2018-01-12T00:00:00",
  "trainStart": "2015-01-16T00:00:00"
}

12 · 2018-07-17T00:00:00 → 2019-01-17T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": -0.38793889532954884,
  "run": 0,
  "testEnd": "2019-01-17T00:00:00",
  "testStart": "2018-07-17T00:00:00",
  "trainEnd": "2018-07-16T00:00:00",
  "trainStart": "2015-07-17T00:00:00"
}

13 · 2019-01-18T00:00:00 → 2019-07-18T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 2.4
  },
  "isSharpe": 0.5729831210160223,
  "run": 2,
  "testEnd": "2019-07-18T00:00:00",
  "testStart": "2019-01-18T00:00:00",
  "trainEnd": "2019-01-17T00:00:00",
  "trainStart": "2016-01-19T00:00:00"
}

14 · 2019-07-19T00:00:00 → 2020-01-17T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 2.4
  },
  "isSharpe": 0.4498765943020883,
  "run": 2,
  "testEnd": "2020-01-17T00:00:00",
  "testStart": "2019-07-19T00:00:00",
  "trainEnd": "2019-07-18T00:00:00",
  "trainStart": "2016-07-19T00:00:00"
}

15 · 2020-01-21T00:00:00 → 2020-07-21T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 2.4
  },
  "isSharpe": 0.556472630848708,
  "run": 2,
  "testEnd": "2020-07-21T00:00:00",
  "testStart": "2020-01-21T00:00:00",
  "trainEnd": "2020-01-17T00:00:00",
  "trainStart": "2017-01-23T00:00:00"
}

16 · 2020-07-22T00:00:00 → 2021-01-22T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": -0.15691434384516234,
  "run": 0,
  "testEnd": "2021-01-22T00:00:00",
  "testStart": "2020-07-22T00:00:00",
  "trainEnd": "2020-07-21T00:00:00",
  "trainStart": "2017-07-24T00:00:00"
}

17 · 2021-01-25T00:00:00 → 2021-07-23T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": 0.10936256514282587,
  "run": 0,
  "testEnd": "2021-07-23T00:00:00",
  "testStart": "2021-01-25T00:00:00",
  "trainEnd": "2021-01-22T00:00:00",
  "trainStart": "2018-01-25T00:00:00"
}

18 · 2021-07-26T00:00:00 → 2022-01-26T00:00:00

{
  "config": {
    "period": 60,
    "std_dev": 1.6
  },
  "isSharpe": 0.3850722808294095,
  "run": 6,
  "testEnd": "2022-01-26T00:00:00",
  "testStart": "2021-07-26T00:00:00",
  "trainEnd": "2021-07-23T00:00:00",
  "trainStart": "2018-07-26T00:00:00"
}

19 · 2022-01-27T00:00:00 → 2022-07-27T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": -0.15043164542544732,
  "run": 0,
  "testEnd": "2022-07-27T00:00:00",
  "testStart": "2022-01-27T00:00:00",
  "trainEnd": "2022-01-26T00:00:00",
  "trainStart": "2019-01-28T00:00:00"
}

20 · 2022-07-28T00:00:00 → 2023-01-27T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": 0.030860824855109822,
  "run": 0,
  "testEnd": "2023-01-27T00:00:00",
  "testStart": "2022-07-28T00:00:00",
  "trainEnd": "2022-07-27T00:00:00",
  "trainStart": "2019-07-29T00:00:00"
}

21 · 2023-01-30T00:00:00 → 2023-07-28T00:00:00

{
  "config": {
    "period": 50,
    "std_dev": 1.6
  },
  "isSharpe": 0.5034340063990634,
  "run": 3,
  "testEnd": "2023-07-28T00:00:00",
  "testStart": "2023-01-30T00:00:00",
  "trainEnd": "2023-01-27T00:00:00",
  "trainStart": "2020-01-30T00:00:00"
}

22 · 2023-07-31T00:00:00 → 2024-01-31T00:00:00

{
  "config": {
    "period": 50,
    "std_dev": 1.6
  },
  "isSharpe": 0.6574714396986291,
  "run": 3,
  "testEnd": "2024-01-31T00:00:00",
  "testStart": "2023-07-31T00:00:00",
  "trainEnd": "2023-07-28T00:00:00",
  "trainStart": "2020-07-31T00:00:00"
}

23 · 2024-02-01T00:00:00 → 2024-08-01T00:00:00

{
  "config": {
    "period": 50,
    "std_dev": 2.0
  },
  "isSharpe": 0.5489178443253857,
  "run": 4,
  "testEnd": "2024-08-01T00:00:00",
  "testStart": "2024-02-01T00:00:00",
  "trainEnd": "2024-01-31T00:00:00",
  "trainStart": "2021-02-01T00:00:00"
}

24 · 2024-08-02T00:00:00 → 2025-01-31T00:00:00

{
  "config": {
    "period": 50,
    "std_dev": 1.6
  },
  "isSharpe": 0.6758271533622142,
  "run": 3,
  "testEnd": "2025-01-31T00:00:00",
  "testStart": "2024-08-02T00:00:00",
  "trainEnd": "2024-08-01T00:00:00",
  "trainStart": "2021-08-02T00:00:00"
}

25 · 2025-02-03T00:00:00 → 2025-08-01T00:00:00

{
  "config": {
    "period": 50,
    "std_dev": 2.0
  },
  "isSharpe": 0.48712484028113895,
  "run": 4,
  "testEnd": "2025-08-01T00:00:00",
  "testStart": "2025-02-03T00:00:00",
  "trainEnd": "2025-01-31T00:00:00",
  "trainStart": "2022-02-03T00:00:00"
}

26 · 2025-08-04T00:00:00 → 2026-02-04T00:00:00

{
  "config": {
    "period": 50,
    "std_dev": 1.6
  },
  "isSharpe": 0.5229873188349763,
  "run": 3,
  "testEnd": "2026-02-04T00:00:00",
  "testStart": "2025-08-04T00:00:00",
  "trainEnd": "2025-08-01T00:00:00",
  "trainStart": "2022-08-04T00:00:00"
}

27 · 2026-02-05T00:00:00 → 2026-08-05T00:00:00

{
  "config": {
    "period": 40,
    "std_dev": 1.6
  },
  "isSharpe": -0.2885678922827421,
  "run": 0,
  "testEnd": "2026-08-05T00:00:00",
  "testStart": "2026-02-05T00:00:00",
  "trainEnd": "2026-02-04T00:00:00",
  "trainStart": "2023-02-06T00:00:00"
}

28 · 2026-08-06T00:00:00 → 2026-11-10T00:00:00

{
  "config": {
    "period": 50,
    "std_dev": 2.4
  },
  "isSharpe": 0.44388245306911334,
  "run": 5,
  "testEnd": "2026-11-10T00:00:00",
  "testStart": "2026-08-06T00:00:00",
  "trainEnd": "2026-08-05T00:00:00",
  "trainStart": "2023-08-07T00:00:00"
}

Run summary

Price observations	4240
Ending equity	0.153801
Total return	-84.62%
Days with target changes	368
Cumulative target change	37000.00%
Cumulative synthetic rebalance turnover	42115.50%
Target gross exposure	0.00%
Target net exposure	0.00%
Target cash residual	100.00%
Effective gross exposure	0.00%
Effective net exposure	0.00%
Effective cash residual	100.00%

Latest asset weights

Market	Target	Effective
UNG	0.00%	0.00%

Target change measures changes in rule weights. Synthetic rebalance turnover also accounts for weight drift after returns; even unchanged targets can require rebalancing. Cash is the signed residual after net exposure.

Accounting changes on identical inputs

{
  "changedReturnDays": 2,
  "correctedTerminalEquity": 0.15380147662386745,
  "legacyTerminalEquity": 0.1604597771101819,
  "maximumDailyReturnDelta": 0.04845813898367868,
  "model": "coherent-lagged-target-v2",
  "terminalEquityDelta": -0.006658300486314439
}

Full run identity and technical record

{
  "inputHash": "b56432154a2f744695540e6876cc21d2074087418661002f3476c46a8d160416",
  "ruleVersion": null,
  "sameInputAccountingCorrection": {
    "changedReturnDays": 2,
    "correctedTerminalEquity": 0.15380147662386745,
    "legacyTerminalEquity": 0.1604597771101819,
    "maximumDailyReturnDelta": 0.04845813898367868,
    "model": "coherent-lagged-target-v2",
    "terminalEquityDelta": -0.006658300486314439
  },
  "sourceHashes": {
    "scripts/build_data.py": "e3dcc988d80a46f7f5a24396bb4ffcac240da8e1249de18a5a282eacccc52320",
    "scripts/enrich_metrics.py": "05611fa9c6de11c261f631cdd8a3314212e7c567f24878a396d057d88752051c",
    "scripts/evidence_v2.py": "c4c6523dbbb9513464dca25150f42cbaac7cbfd71bd8afdc855659b2c763d72e",
    "scripts/performance_io.py": "42282ff10697ceb56dc7ca2edf79538c2e1e8344da8e6d69ceac4d16ba5c668a",
    "scripts/recompute_evidence.py": "a7f59004cb7c356940ee3748439024cd83cac434acb90c0bead2ae19f81be127",
    "scripts/replay_evidence.py": "4e04d937c19b9040da6dc9c3c80ac22181b45ffba0131f3a63a4def95865fad6",
    "scripts/requirements.txt": "7424048072a2cf3d8c3b4e4f37a5b36c1e7ff3732d893c43295d17d85c5f97d1",
    "scripts/strategies/linear_regression_channel.py": "aa462f6f30a5d16446fee606a47e8a92c34725d585bcaf776e75f7917e4c3645",
    "scripts/strategy_evidence.py": "a3a93d62a84298470af3eb8e7d74e9f0ca0d29cf672d1d2d5449b41904e3d58c"
  },
  "runId": "50e579e8269be60456602302f86093b53f5a752926091fe6539db646ced04371",
  "generatedAt": "2026-11-11T10:55:31.449953+00:00",
  "executionModel": {
    "cash": "1 - sum(weights); residual may exceed one or be negative for shorts/leverage.",
    "decision": "Target weights use bars through close t; previous targets earn (t-1,t].",
    "feesBasisPoints": 0,
    "fills": "Synthetic daily target allocations, not broker orders or fills; financing/borrow costs omitted.",
    "id": "synthetic-daily-target-v3",
    "rebalanceTurnover": "Sum(abs(target[t] - target[t-1]*(1+assetReturn[t])/(1+grossReturn[t]))); one-way risky notional at close t after returns, divided by pre-trade equity.",
    "return": "Sum of previous target weight times each asset close-to-close return; cash return is zero.",
    "targetChange": "Sum(abs(target[t] - target[t-1])); changes in rule targets only.",
    "wfo": "Candidate retains its own hypothetical state; first test return uses previous published target."
  }
}

Downloads

1. Download a complete bundle

2. Replay the arithmetic

3. Obtain and compare inputs

4. Recompute the original strategy

Parameters used

Walk-forward selections

Run summary

Latest asset weights

Accounting changes on identical inputs

Get the weekly Arena report